The Affect of Exercise Past, Present, and Future: The Role of Affective Memory and Affective Forecast in the Affective Response-PA Behaviour Relation

Affective responses experienced during exercise have predicted future physical activity (PA) behaviour (Williams, Dunsiger, Jennings & Marcus, 2012). When applied to exercise, the premise of the affect heuristic (Slovic Finucane, Peters & MacGregor, 2002) is that decision making is influenced by the memory of the affective response experienced from exercise. However, affective memories have not predicted PA (Hargreaves & Stych, 2013). Alternatively, affective memories may predict affective forecasts, which have been suggested as a predictor of PA behaviour (Williams, Anderson & Winett, 2005), but there has been no direct test of the relations between affective memories from exercise, affective forecasts for exercise, and PA. Therefore, the purpose of this thesis was to investigate the role of affective memories from exercise and the affective forecasts for exercise in the affective response-PA behaviour relation. The first study was designed to test the relations between affective forecasts and affective responses during exercise, and the relations between affective memories and affective forecasts for subsequent exercise. Twenty females (M = 39.4 years, SD = 11.2) participated in two identical moderate intensity exercise sessions one week apart, followed by a graded exercise test. The significant predictive relation between affective forecast measured before the first exercise session and the mean affective response during exercise session was stronger for the second exercise session (explaining 60% variance), compared to the first exercise session (explaining 29% variance). The affective forecasts were significantly more positive for the second exercise session (M = 3.6, SD = 3.4) compared to the first exercise session (M = 0.9, SD = 3.9). There were no differences in affective forecasts for those who were reminded about their previous exercise experience when making an affective forecast compared to those who were not. Affective memories measured 15 min, two days, and seven days after exercise significantly predicted the affective forecast for the second exercise session explaining 54%, 52%, and 76% variance respectively. These results indicate that if affective responses to exercise are positive, then subsequent affective forecasts are more likely to be positive due to the positive affective memories from that previous exercise experience. The second study was designed to test whether affective forecasts mediated the relation between affective memories and PA behaviour over time. Twenty-nine females and two males (M = 38.2 years, SD = 8.8) participated in a graded exercise test, which was followed two days later by an exercise session set at an intensity around ventilatory threshold. Affective memories from the exercise session predicted affective forecasts for a subsequent exercise experience after 15 min, one week, one month, and three months, significantly explaining 74%, 34%, 42%, and 39% of the variance respectively. Neither affective forecasts nor affective memories predicted PA behaviour measured one week, one month, or three months after the exercise session. Therefore, the conditions necessary to test for the mediation were not met. Despite, the strong theoretical rationale that affective forecasts would predict PA behaviour, the results did not support this conclusion. However, limitations with respect to sample size and the method of measuring affective memory may explain these surprising findings. The third study was a qualitative design to explore what factors people thought about when creating their affective forecasts for exercise. The thirty-one participants from study two were asked to explain what they were thinking about to create their affective forecast for exercise 15 min, one week, one month, and three months after the exercise session in the laboratory. A thematic analysis was applied to the transcribed semi-structured interview data. Four higher-order themes were identified as the aspects of exercise that influenced the creation of their affective forecasts, these were: the Interpretations of the Exercise Intensity, the Outcomes from Exercise, the Exercise Context, and the Enjoyment from Exercise. Collectively, the results demonstrated an almost cyclical relation between affective memories, affective forecasts, and affective responses in the exercise context. The last piece of the puzzle is for future research to investigate the relations between these affective facets and PA behaviour